Copyright protection of digital images by embedded unperceivable marks

摘要

Digital watermarking has been proposed as a solution to the problem of copyright protection of multimedia documents in a networked environment. It makes it possible to tightly associate a code to an image, allowing the identification of the data creator, owner and so on. In this paper, after a review of some important watermarking techniques, a new watermarking algorithm for digital images is presented: the method, which enhances a previous technique proposed by Cox et al. [1] [I.J. Cox, J. Kilian, T. Leighton, T. Shamoon, Secure spread spectrum watermarking for multimedia, NEC Research Institute Technical Report 95-10], operates in the frequency domain, by embedding in a selected set of DCT coefficients a sequence of randomly generated real numbers. By exploiting their statistical characteristics, it is possible to reliably extract the embedded sequences without resorting to the original unmarked image. Experimental results demonstrate that the watermark is robust to several image transformations.